预测概率 [英] Predicting probability

问题描述

尝试使用sklearn中的SVC进行分类问题.给定一堆数据，并且信息告诉我某个学科是否在某个班级中，我希望能够给出一个新的未知学科在班级中的可能性.

Trying to use SVC from sklearn to do a classification problem. Given a bunch of data, and information telling me whether some subject is in a certain class or not, I want to be able to give a probability that a new, unknown subject is in a class.

我只有2个课，所以问题是二进制的.这是我的代码和一些错误

I only have 2 classes, so the problem is binary. Here is my code and some of my errors

from sklearn.svm import SVC
clf=SVC()

clf=clf.fit(X,Y)


SVC(probability=True)
print clf.predict_proba(W) #Error is here

但是它返回以下错误:

NotImplementedError: probability estimates must be enabled to use this method 

我该如何解决?

推荐答案

您必须使用probability=True

from sklearn.svm import SVC
clf=SVC(probability=True)
clf.fit(X,Y)
print clf.predict_proba(W) #No error

您的代码使用概率估计值创建了一个SVC并将其丢弃(因为您没有将其存储在任何变量中)，并使用了存储在clf中的先前的SVC(无概率)

Your code creates a SVC with probability estimates and discards it (as you do not store it in any variable) and use some previous SVC stored in clf (without probability)

这篇关于预测概率的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！